SpaceXSI operates multi-campus, hyperscale GPU infrastructure in Memphis, where every minute of downtime costs training throughput and customer commitments. The Operations Engineering team owns 100% GPU uptime, end to end, 24/7. This is a new engineering and specialist team serving both Site Operations (compute) and Facility Operations (power and cooling). It owns incident response and the engineering behind the response, holds central control authority over what compute is impacted or drained, drives day-to-day sustainment of live datahalls, and leads the deployment engineering that brings new datahalls online. The team will also build the automation behind our NOC Bot, the automated monitoring layer that replaces a traditional technician NOC. Think of it as the datacenter equivalent of SpaceX Launch Engineers and Launch Controllers: engineers and specialists with system-level knowledge of the whole site.

Staff a 24/7 rotation that owns incident management end to end, from customer impact (internal Grok teams and external datahall tenants) back to the facility.

Hold central control authority: determine what compute is impacted or drained and own the system-wide response.

Own day-to-day site sustainment operations, including resolving issues impacting rack-level compute.

Disposition daily issues across both Site Ops (compute) and Fac Ops (power and cooling).

Lead acceptance and deployment testing campaigns to bring new data halls fully online.

Develop production scripts and automation for new datahall bring-up, and build the automation behind the NOC Bot.

Develop and maintain manufacturing-planning-style work instructions so technicians work more efficiently and safely.

Own and document the live, changing state of datahall operation as new halls come online.

Enable Site Ops and Fac Ops technicians to do their jobs more efficiently and safely to increase GPU uptime.

Supercompute Operations Specialist - Memphis

Responsibilities

Qualification

Required

Preferred

Company

xAI

More roles at xAI